n8n RAG Chatbot

Workflow-Powered RAG with n8n + pgvector + OpenAI

Documents (0)

Upload .txt files to build your knowledge base

Welcome to n8n RAG Chatbot!

Upload documents and ask questions about them.

Powered by n8n workflows

Knowledge Base

Upload documents to build your custom knowledge base

Smart Search

Vector similarity search finds relevant context

AI Responses

GPT-4o generates answers based on your data

Tech Stack & Architecture

n8n Workflows

  • Document Ingestion - Webhook-triggered text processing
  • Query Processing - Vector search + GPT-4o response
  • OpenAI Integration - text-embedding-3-small (1536 dims)
  • PostgreSQL + pgvector - Vector storage with HNSW indexing

Frontend

  • Next.js 15 - React framework with App Router
  • Tailwind CSS - Glassmorphism styling
  • Direct Webhook Calls - No backend API needed
  • Session Privacy - localStorage-based isolation

n8n Workflow Data Flow

Document Ingestion Workflow:

Webhook Trigger
Validate Input
Create Document
Chunk Text
OpenAI Embeddings
Store in Postgres

Query Workflow:

Webhook Trigger
Embed Question
pgvector Search
Format Context
GPT-4o Response
JSON Response

Key Technical Features

No Backend Code: n8n handles all processing
Semantic Chunking: 500 tokens with 50-token overlap
Visual Workflow: Easy to modify and extend
Source Citations: Track which documents answered questions